Sync meeting on EESSI test suite (2024 03 14)

Planning

• every 2 weeks on Thursday at 14:00 CE(S)T
• next meetings:
  • 28th March, 2024.
  • 18th April, 2024.
  • 2nd May, 2024

---

Meeting (2024-03-14)

• Disable running CUDA-capable modules on CPU #110

  • Merged
  • Now, for Snellius GPU config here, it will NOT run OSU on host buffers. I don't think this is desired, I think we should have OSU not use the hook... → (Satish)

• This was merged, updates in CI configs with latest version of EESSI and some polishing in the scripts.

  • Found that, on the AWS cluster, the number of OSU iterations are too many (1000) should be changed to (20) and also reduce the warm-up iterations to 5 from 10. → (Satish)

• 0.2.0 is released!! Release notes → here

• Create release notes for 0.2.0 (Lara) → ☑

• Sam: Cleanup! :) → ☑

• Add ALWAYS_REQUEST_GPU to surf_snellius.py config (Caspar) → ☑

• Add CUDA-capable OSU to EESSI (Caspar) → ☑

  • LMOD hook was required which caused some problems.

• Add the --report-file option to daily runs on Vega / Karolina, or put it in the config file (?) (Caspar) → ☑

Next steps?

• Satish: OpenFOAM test

  • Example in the OpenFOAM demo is not part of OpenFOAM 11 anymore, but we could still copy it from an older version and ship it with the test suite
  • software-layer demo is also broken for this reason
  • Formatting has changed for Form dictionary in OpenFOAM 11 compared to previous versions
  • After formatting adaptations: getting segfaults for the software-layer demo (motorbike case)
  • 2 test cases at least, motorbike takes a bit long.
  • Lid driven cavity.
  • Split the test cases into many steps.
  • Test on the 2 forks of OpenFOAM.
  • Need to really check the starting scale for motorbike test.
  • Useful to have both fast and long tests.

• ESPResSo

  • Satish can help Xin out.

• PyTorch (Caspar)

  • Basic CNN -> pure PyTorch.
  • Torchvision for loading other standard models into PyTorch.
  • Microsoft Super benchmarks?

• LAMMPS (Lara)

  • made a script for LJ and rhodo.
  • Create a PR for the job scripts and run scripts.

• Test json-export format from ReFrame → (Sam)

• CP2K (Sam)

• Create a "How to contribute to the test-suite?" in EESSI documentation.

  • Few examples, one with MPI.